Training Data & Original Data for Autonomous Excavation

Citation Author(s):: JAEMYUNG HUH (KOREA UNIVERSITY)

DAEHIE HONG (KOREA UNIVERSITY)
Submitted by:: JAEMYUNG HUH
Last updated:: Thu, 11/03/2022 - 09:24
DOI:: 10.21227/s29n-y527
Data Format:: *.csv;*.ply;

161 views

Categories:

Artificial Intelligence

Keywords:

LSTM-RNN

ACCESS DATASET CITE

Abstract

The data in this file include not only original data which is not preprocessed, but also training data to train our deep learning : bucket tip trajectory planning AI.

These csv files includes pressure, position, and angle of bucket tip, and current time, which were measured in real time excavation. Also, the point cloud is captured before each excavation.

Instructions:

It will be helpful when you use these data with pandas module from python.

I use pandas to preprocess the original data. Especially, the point cloud is too big to extract features. So, I convert point clouds to just polynomials, which represent terrain as mathematical expression.

Funding Agency

Ministry of Trade, Industry and Energy(MOTIE, Korea)

Grant Number

20010776